sampling testing

Hi all,

This sample testing question has been bugging me for a long time at work. I would appreciate it if someone could enlighten me.

I have a population consisting of 9000 people who are classified into 3 groups:
A = 2000
B = 4000
C = 3000

I randomly create a sample consisting of 3000 observations:
A = 800
B = 1400
c = 800

What's the best way to test the likelihood of obtaining such a sample and whether this sample is indeed random.